Opening the black box of accuracy: an analysis of interactions among errors in costing systems

Accuracy of costing systems is a central issue in management accounting research. Previousresearch on accuracy has identified three main sources of errors affecting the design of a two stagecosting system such as Activity Based Costing (ABC): aggregation, specification and measurementerrors. Aggregation errors are modelled as a reduction in the number of resource or activity poolsused to allocate resource costs to cost objects, and this modelling implies that biased and benchmarkallocation mechanisms are not directly comparable. As a consequence, the interaction among errorsis analysed in terms of cost figures resulting from the biased versus the benchmark allocationmechanism, instead of comparing the allocation mechanisms step by step. In this sense, accuracy isa black-box that needs to be opened to fully understand the mechanisms underlying costing errorsand their interactions. In this paper, we propose an analytical model in which aggregation errors areintroduced without reducing the number of resources and activity pools. This approach allows for adirect comparison of the biased and the benchmark systems, and allows us to deepen our knowledgeon error interactions. This approach allows us to understand how accuracy is affected by the errorstructure and by the resource consumption pattern. Furthermore, the effect and the interaction ofdifferent kinds of error (aggregation, specification and measurement) is analysed.